Bipolar disorder (BD) is the Axis I psychiatric condition most strongly associated with substance use disorder (SUD); diagnostic co-occurrence is particularly high between BD and alcohol use disorder (AUD). Individuals with co-occurring SUD and BD (SUD+BD) have substantially worse clinical outcomes than those with either BD or SUD alone. Nonetheless, little is known about optimal treatment for individuals with SUD+BD; response to lithium appears to be poor, and only one double-blind, randomized, placebo-controlled trial of valproate has demonstrated improved drinking outcomes in this population. Traditionally, treatment trials for SUD+BD have investigated medications that have been FDA approved to treat either BD or SUD in hopes that such medications would prove efficacious in individuals with SUD+BD. A different approach to selecting, and ideally developing, medications for SUD+BD treatment trials would be to target neurochemical dysfunctions characteristic of individuals with both BD and SUD. Our lab recently demonstrated unique disturbances in prefrontal gamma- Aminobutyric acid (GABA) and glutamate concentrations in this population using proton magnetic resonance spectroscopy (1H-MRS), with individuals with co-occurring alcohol dependence (AD) and BD having significantly lower levels of GABA and glutamate relative to individuals with BD alone, AD alone, or healthy controls. Lower levels of prefrontal GABA and glutamate were in turn associated with elevated impulsivity and alcohol craving. The proposed 3-week, double-blind, crossover, proof of concept study will evaluate: a) whether medications that have been demonstrated to normalize cortical GABA (i.e., gabapentin) and glutamate (i.e., N-Acetylcysteine [NAC]) concentrations in individuals with epilepsy and cocaine dependence, respectively, may similarly act to normalize prefrontal GABA and glutamate levels in individuals with AUD+BD, and b) whether normalization of prefrontal GABA and glutamate levels will be associated with improvements in functional brain activity to tasks that assess core neurobehavioral deficits of AUD and BD (i.e., response inhibition, alcohol cue-reactivity), as well as drinking and mood symptoms. Positive results may support investigation of gabapentin and/or NAC as adjunctive treatments for AUD+BD in large-scale, randomized clinical trials. Most importantly, the proposed study may provide successful demonstration of a neuro-behavioral, multimodal neuroimaging platform for evaluating the potential promise of GABAergic and glutamatergic drugs for AUD and/or BD, as well as other conditions marked by GABAergic/glutamatergic dysfunction.